High-power test of an interdigital $H$-mode drift tube linac for the J-PARC muon $g-2$ and electric dipole moment experiment

We conducted a high-power test of a prototype cavity of a 324-MHz interdigital H-mode drift tube linac (IH-DTL) for the precise measurement of the muon anomalous magnetic moment ($g-2$) and electric dipole moment (EDM). This prototype cavity (short IH) was developed to verify the fabrication methodology for the IH-DTL cavity with a monolithic drift tube structure. The electromagnetic field distribution was measured and compared with the finite element method simulation results, and the fabrication accuracy of the monolithic drift tube was confirmed to satisfy the requirements. After 40 h of conditioning, the short IH was stably operated with an rf power of 88 kW, which corresponds to a 10% higher accelerating field than the design field ($E_0$) of $3.0\mathrm{MV}/\mathrm{m}$. In addition, the thermal characteristics and frequency response were measured, verifying that the experimental data were consistent with the three-dimensional model. In this paper, the design, fabrication, and low-power and high-power tests of this IH-DTL for muon acceleration are described.

Figure 1

Optimized phase array as a function of the cell number [16]. The short IH corresponds to the first six cells of the full IH.

Figure 2

Mechanical structure of the short IH.

Figure 3

Side view of the center plate.

Figure 4

Simulated on-axis EM fields of the short IH at the nominal voltage. The average accelerating field is $3.0\mathrm{MV}/\mathrm{m}$.

Figure 5

Calculated phase-space distributions at the short-IH exit. (a) the horizontal divergence angle $x^{\prime }$ vs $x$, (b) the vertical divergence angle $y^{\prime }$ vs $y$, (c) energy spread $\mathrm{\Delta }W(W-1.3\mathrm{MeV})$ vs phase spread $\mathrm{\Delta }\varphi$, and (d) $y$ vs $x$.

Figure 6

Fabricated center plate of the short IH.

Figure 7

Cross-sectional view of the short IH. The tuners have no rf contactors to simplify the structure.

Figure 8

Top: frequency shift as functions of the tuner insertion lengths of each tuner. The tunable ranges with the three tuners were 3.24 and 3.27 MHz in the measurement and simulation, respectively. (bottom) The unloaded $Q$ ($Q_0$) shift dependence on $L_{\text{tuner}}$.

Figure 9

Frequency shift distribution obtained with the bead pull measurement. Top: the measured and simulated frequency shift. The gray bands show the gap regions. Bottom: the difference between the measurement and the simulation (field error) in the gap region.

Figure 10

Photograph of the experimental apparatus of the high-power test of the short IH.

Figure 11

Block diagram of the high-power test system. The interlock is triggered by rf powers and vacuum levels.

Figure 12

Conditioning history of the short IH. The red arrows in the top plot indicate the cavity trip caused at the nominal duty factor. The temperature data show the difference between the average of the three thermocouples and the ambient temperature. The surface temperature of the cavity was measured at three different locations and showed no dependence on location.

Figure 13

rf waveforms of forward (top), reflected (middle), and cavity pickup voltages (bottom) at the nominal peak power of 75 kW.

Figure 14

Simulated power dissipation distribution on the cavity surface with the peak power of 88 kW and nominal duty factor.

Figure 15

Steady-state temperature distribution of the short IH with the peak power of 88 kW and a nominal duty factor. The distribution shows the temperature difference from the initial state of a uniform 24°C.

Figure 16

Top: frequency shifts as functions of time after the rf power of 88 kW was turned on. The solid lines show the fitting results. Bottom: temperature drifts as functions of time. The data were plotted with the difference between ambient and cavity surface temperatures. The DT’s temperature was increased rapidly within 0.3 h in the simulation.